The histological characteristics, age-related thickness change of skin, and expression of the HSPs in the skin during hair cycle in yak (Bos grunniens)

Objective This experiment was conducted to study the histological characteristics, age-related thickness changes, and expression of HSPs in the skin of yak. Methods A total of 20 yaks (10 males and 10 females) were used. Different regions of the normal skin of three different ages (newborn, half-year-old and adult) of yaks were harvested for histological study and thickness measurement. Biopsy samples were taken from the scapula regions of the skin from the same five approximately 1-year-old yaks during the hair cycle (telogen, anagen and catagen). RT-PCR, western blot and immunohistochemistry methods using the mRNA and protein levels were used to detect the expression of HSP27, HSP70 and HSP90. RT-PCR method was used to detect the mRNA expression of CGI-58 and KDF1. The IPP6.0 software was used to analyze the immunohistochemistry and measure the thickness of the skin. Results The general histological structure of hairy yak skin was similar to other domestic mammals. The unique features included prominent cutaneous vascular plexuses, underdeveloped sweat glands, a large number of nasolabial glands in the nasolabial plate, and hair follicle groups composed of one primary follicle and several secondary follicles. The skin, epidermis and dermis thickness did vary significantly between different body regions and different ages. The thickness of the skin, epidermis and dermis increased from newborn to adult in yaks. Yak skin thickness decreased from dorsally to ventrally on the trunk. The skin on the lateral surface was thicker than the skin on the medial surface on the limbs. HSP27, HSP70 and HSP90 showed different expression patterns during the hair cycle using RT-PCR, western blot and immunohistochemistry methods. The expression of HSP27 mRNA and protein in the anagen stage was the highest, followed by the catagen stage, and the expression in the telogen stage was the lowest. The expression of HSP70 mRNA and protein in the telogen stage was the highest, followed by the anagen stage, and the expression in the catagen stage was the lowest. The expression of HSP90 mRNA and protein in the anagen stage was the highest, followed by the telogen stage, and the expression in the catagen stage was the lowest. HSPs were mainly expressed in the outer root sheath of hair follicle during the hair cycle, also expressed in epidermis, sebaceous gland and sweat gland in the skin of Yak. The expression of CGI-58 mRNA in the anagen stage was the highest, followed by the catagen stage, and the expression in the telogen stage was the lowest. The expression of KDF1 mRNA in the telogen stage was the highest, followed by the catagen stage, and the expression in the anagen stage was the lowest. Meaning In this study, we examined and fully described the histology of normal skin in Yak and measured the skin thickness of different ages and different regions in Yak. These data may be useful to better understand and appreciate the adaptability features of yak skin. Our investigation reports the expression patterns of HSPs in yak skin for the first time. The different expression pattern of HSPs during the hair cycle suggests they may play different roles in yak hair follicle biology.

Introduction Yak (Bos grunniens) is a special plateau mammal that lives in the extreme environments of the Tibetan highlands, which has the basic features of extreme cold, high altitudes with reduced oxygen content in the air, and high ultraviolet radiation. The altitude where yaks live normally is over 3000-6000 meters high. The annual temperature of this area is -3 to 3˚C, and the extreme lowest temperature is -40˚C. Yaks play an important role in normal life for people living in the plateau, such as providing meat, milk and wool, packing goods and materials and riding.
The skin is the largest organ of the body and serves many functions, such as protection against environmental aggressions (cold, intense radiation and sandstorms), sensation, metabolism and thermoregulation. There are some studies of the skin histology of some mammals such as llamas [17], sheep [18], ferrets [19] and camels (Camelus dromedaries) [20]. It is an accepted fact that skin varies considerably in thickness based on its site and age. In 2009, Volkering measured the skin thickness over the equine body [21]. Many of the related works that have been conducted on different body regions and ages have been done in humans [22][23][24]. Only a few studies have been published on the histological research and measurement of yak skin, but there are no detailed data for skin histologic characteristics, thickness changes or the relationship between structure and adaptability in yak.
Heat shock protein (HSP) is one type of molecular chaperone and includes five major groups: 20-30, 60, 70, 90 and 110 kDa based on molecular size [25]. HSPs are involved in protein folding, assembly, transport and regulation of cell growth and differentiation [26]. Heat shock protein-27 (HSP27) is a member of the small heat shock proteins (sHSP). The primary structure of HSP27 is highly homologous to other members of the sHSP family; it contains the conserved α-crystallin domain and differs in the C-and N-terminal regions. HSP27 is expressed in all human tissues, including astrocytes and primary neuronal cells, but is mainly found in skeletal, smooth and cardiac muscles [27]. HSP27 protein is expressed in a differentiation-related pattern [28][29][30][31][32]. For example, keratinocytes of the upper epidermis express higher levels of HSP27 than basal cell keratinocytes in normal human skin. In developing human skin, HSP27 protein expression correlates with increasing epidermal differentiation and trichilemmal keratinization [33].
HSP70 included two major proteins: constitutively expressed HSC70 and stress-inducible HSP72 [34]. HSC70 is expressed in practically all organs and tissues and functions as ATPdependent molecular chaperone under normal conditions [35][36][37]. HSP70 plays an important role in cell apoptosis through its ability to inhibit apoptosis. HSP70 can regulate cell apoptosis at different levels such as affecting some transcription factors involved in the expression of Bcl-2 family [38].
HSP90 belongs to another important HSPs family. It is a kind of abundant protein expressed in all eukaryotic calls [39,40]. HSP90 is highly conserved and also is an ATPdependent chaperone. HSP90 can maturate, stabilize and activate a range of client proteins through form complex [40,41]. Many of these client proteins are involved in cell growth, proliferation and survival.
Hair follicle (HF) cycling transitions include telagen, anagen and catagen. These phases are controlled by molecular switches such as HSPs. To date, nothing is known about HSPs protein expressions in the skin of yaks or their relation to the cycling changes in HF. In this investigation, we tested the expression patterns of HSP27, HSP70 and HSP90 in normal yak skin during their hair cycle.

Experimental animals and treatments
A total of 20 yaks (10 males and 10 females) from the Gannan Tibetan Autonomous Prefecture in Gansu province and Xining City in Qinghai province that were humanely euthanized for reasons unrelated to the skin were used in this study. Yaks were purchased from Jianguo Ma and Ming Liu, the small holders in Gannan Tibetan Autonomous Prefecture of Gansu Province and Datong County of Qinghai Province (China). Yaks were permitted as experimental animals by the owners. In order to maintain the original habitat, the yaks were executed and samples were collected in the local instead of being housed at the university. All of the yaks were in good nutritional condition and were distributed evenly into four groups (newborn, half-year-old, 1-year-old and adult) ( Table 1). In this study, the experimental animals were all handled according to the Animal Ethics Procedures and Guidelines of the People's Republic of China, and the study was approved by the Animal Ethics Committee of Gansu Agricultural University.
Twenty-five different regions (Fig 1) of the normal skin of three different age (newborn, half-year-old and adult) yaks were harvested for histological study and thickness measurements. All of the animals were euthanized by intravenous injection of pentobarbital sodium (150 mg/kg body weight) for animal welfare and safety of experimenter.
Five of the 1-year-old yaks that were used for the hair follicle study were kept under the same natural photoperiod and temperature conditions. Biopsy samples of the skin during the hair cycle (telogen, anagen and catagen) were taken from the scapula region. Skin specimens used for immunohistochemistry were stored in 4% paraformaldehyde solution, and skin specimens used for RT-PCR and WB were stored at -80˚C.

Measurement and data analysis
Skin thickness was measured using a micrometer. The thickness of the epidermis and dermis was measured using a computerized light microscope (Olympus DP71) and morphometric software (Image-Pro plus 6.0). The epidermis was measured from the free margin of skin to the dermis papillae and epidermis ridge. The dermis was measured in the same way from the epidermis ridge and dermis papillae to the dermal-fat junction [24]. Data were expressed as the mean ± standard deviation (SD). Statistical analysis was performed using the Statistical Package for Social Science software, version 19.0 (SPSS Inc., Chicago, IL, USA). Statistical analysis was primarily conducted using a one-way analysis of variance (ANOVA). A P value of P<0.05 was considered statistically significant.

Histologic characteristics of skin structure
Epidermis. There was no obvious epidermal interpapillary peg in yak skin, but the epidermis appeared undulating from the epidermis of the opening in the hair follicle down to the dermis (Fig 2). The epidermis of hairy skin in yak consisted of four layers: stratum corneum, stratum granulosum, stratum spinosum and stratum basale, whereas the glabrous skin also included stratum lucidum, such as the nasolabial plate and the hooves.
The stratum corneum was made of flattened, anucleated, scale-like cells that were fibrous and could easily fall off. The stratum lucidum was a pink uniform band that was found in the planum nasolabiale and was hoof coronal and hoof sphere. The stratum granulosum consisted of one layer of flattened cells, which would be keratinized into the corneum. The depth of the cell layer in the stratum spinosum was varied by age changing. In newborn yak, there were only 1-2 layers, although it increased to 3-4 layers in the half-year-old and adult yak. The stratum basale was made of cubical cells with large nuclei lying perpendicular to the basical membrane (Fig 3A and 3B).  Dermis. The dermis contained the papillary, which was in contact with the epidermis, and reticular layers, which were in contact with the underlying hypodermis. The dermis consisted of numerous fiber types and few cell types and contained numerous blood and lymphatic vessels, nerves, arrector pili, sweat glands, sebaceous glands, hair and hair follicles (Fig 2).
The dermis contained a variable amount of collagen and a few elastic and reticular fibers, which often were intertwined with each other (Fig 4A and 4B). The collagen fibers were arranged in bundles and appeared blue-green using Masson's trichrome staining ( Fig 4A). Compared with the papillary layer, collagen fiber bundles were thicker in the reticular layer. Elastic fibers were primarily located in the papillary dermis and surrounding vessels, which appeared black-green when using WVG staining ( Fig 4B). In the hoof, the epidermal lamellae and the dermal lamellae fit together to form a tight junction ( Fig 4C). A large number of capillary plexuses was distributed around the boundary between the papillary and reticular layers ( Fig 4D). The arrector pili was a thin smooth muscle bundle that was located in the papillary layer and extended up to the epidermis (Fig 2). There was also some adipose tissue in the deep dermis in some body regions such as the hoof sphere ( Fig 4E). In yak, few sweat glands were located in the deep dermis, and the secretion was blue-violet by AB-PAS staining (Fig 4F), which meant that the secretion was an acidic-neutral complex that contained many saccharides. Many nasolabial glands were distributed in the dermis of the nasolabial plate, which belonged to the branched tubuloacinar gland (Fig 4G). The sebaceous glands were well developed and distributed around the hair follicles ( Fig 4H).
Hair follicle. The yak hair was longer and denser than other cattle breeds, and the thickness was different, with hair follicles that were variable in size, were evenly distributed in the papillary and upper reticular layers and often formed in groups. One hair follicle group consisted of one primary follicle (PF) and several secondary follicles (SF), which were accompanied by sebaceous glands. The hair follicle group was surrounded by a connective tissue sheath (CTS) (Fig 5A). The hair follicle was composed of the dermal root sheath (DRS), outer root sheath (ORS) and inner root sheath (IRS) (Fig 5B). The dermal root sheath was made of connective tissue. The outer root sheath possessed 3-4 layers, whereas the inner root sheath was composed of three layers: Henle's layer, Huxley's layer, and an internal cuticle. Specifically, Henle's layer consisted of one single layer of cubical cells with flattened nuclei, and Huxley's layer was composed of a cell layer with flattened nuclei. By contrast, the internal cuticle had one cell layer, which had been keratinized.

Measurement of skin thickness
In the newborn group, the skin thickness ranged from 624 to 1538 μm; the inguinal region was thinnest (624 μm), and the back and cheek were the thickest (1500-1538 μm) (Tables 3 and 4). The thickness of the epidermis varied from 16.073 to 29.307 μm. The thickness in the lateral crus, cheek and inguinal regions ranged from 16.073 to 17.104 μm. The back, dorsal neck and forehead were relatively thick (21.913-23.584 μm), and the metacarpus was thickest (29.307 μm) (Tables 3 and 4). The thickness of the dermis varied from 389.956 to 948.520 μm, and that in the metatarsus, waist, axilla, lateral crus and the inguinal region ranged from 389.956 to 457.078 μm. The abdomen, dorsal neck and buttock were relatively thick (545.228-683.58 μm), and the back was thickest (948.520 μm) (Tables 3 and 4).
In the half-year-old group, the skin thickness ranged from 976 to 2364 μm; the inguinal region was thinnest (976 μm), and the cheek, ventral neck and dorsal neck were the thickest (2264-2364 μm) (Tables 3 and 4). The thickness of the epidermis varied from 22.458 to 63.594 μm. The thickness in the costal region, brachia, thorax and abdomen ranged from 22.458 to 25.973 μm. The metatarsus, buttock and dorsal neck were relatively thick (30.194-34.372 μm), and the metacarpus was thickest (63.594 μm) (Tables 3 and 4). The thickness of the dermis varied from 507.742 to 1210.813 μm. The thickness in the inguinal region, axilla and medial region of the forearm ranged from 507.742 to 581.731 μm. The lateral neck, cheek, costal region and thigh were relatively thick (687.223-903.122 μm), and the back was thickest (1210.813 μm) (Tables 3 and 4).
In the adult group, the skin thickness ranged from 2052 to 5934 μm; the axilla and inguinal regions were thinnest (2052 μm, 2190 μm), and the back and metacarpus were thickest (5934 μm, 5324 μm) (Tables 3 and 4). The thickness of the epidermis varied from 34.211 to  Characteristic, thickness change and expression of HSPs in yak skin  (Tables 3 and 4). The epidermis accounted for 2.258-5.879% of the entire skin in the newborn group, 2.647-5.432% in the half-year-old group, and 2.05-5.627% in the adult group (Table 3).
The thicknesses of the epidermis and dermis increased with age from newborn to adult. The age-related thickness changes differed significantly in the newborn, half-year-old and adult groups ( Table 5). The differences in thicknesses of both the dermis and the skin were Characteristic, thickness change and expression of HSPs in yak skin statistically significant among the three age groups (P<0.05), whereas the thicknesses of the epidermis varied among the three groups. In the thorax, back, costal region and abdomen, the thickness of the epidermis showed no significant difference between the newborn and halfyear-old group (P>0.05) but was significantly different compared to the adult group (P<0.05). A schema graph was provided to show the changes in adult yak according to the data we obtained (Fig 6).

Expression of HSP27 in skin during hair cycle
A fluctuation in the relative expression levels of HSP27 mRNA during the hair cycle was shown in Fig 7A. In the hair cycle, the highest level of HSP27 mRNA expression was found during the anagen stage, whereas the lowest expression level was found in the telogen stage. The expression level in the catagen stage was between the anagen and telogen stages. The expression level showed significant difference between anagen and telogen stages (P<0.05) as well as between anagen and catagen stages (P<0.05), but there was no difference between telogen and catagen stages (P>0.05). A similar expression pattern was observed for the HSP27 protein using western blot analysis (Fig 8A and 8B). The highest expression level was seen in the anagen stage, followed by the catagen stage, with the lowest level seen in the telogen stage. The result showed that the HSP27 protein expression levels were significantly different among hair cycle (P<0.05). HSP27 was mainly expressed in the outer root sheath of the secondary follicle during the hair cycle, also expressed in epidermis and sebaceous gland in the skin of yak (Fig 9). Characteristic, thickness change and expression of HSPs in yak skin Expression of HSP70 in skin during hair cycle A fluctuation in the relative expression levels of HSP70 mRNA during the hair cycle was shown in Fig 7B. In the hair cycle, the highest level of HSP70 mRNA expression was found during the telogen stage, whereas the lowest expression level was found in the catagen stage. The expression level in the anagen stage was between the telogen and catagen stages. There was no significant difference among three stages (P>0.05). A similar expression pattern was observed for the HSP70 protein using western blot analysis (Fig 8A and 8C). The highest expression level was seen in the telogen stage, followed by the anagen stage, with the lowest level seen in the catagen stage. However, there was significant difference of HSP70 protein expression levels among three stages (P<0.05). HSP70 protein expression was observed in the epidermis, sebaceous gland, sweat gland and outer root sheath of hair follicle in the skin (Fig 10).

Expression of HSP90 in skin during hair cycle
A fluctuation in the relative expression levels of HSP90 mRNA during the hair cycle was shown in Fig 7C. In the hair cycle, the highest level of HSP90 mRNA expression was found during the anagen stage, whereas the lowest expression level was found in the catagen stage. The expression level in the telogen stage was between the anagen and catagen stages. There was no significant difference among three stages (P>0.05). A similar expression pattern was observed for the HSP90 protein using western blot analysis (Fig 8A and 8D). The highest expression level was seen in the anagen stage, followed by the telogen stage, with the lowest level seen in the catagen stage. The expression levels of HSP90 protein showed significant difference between anagen and telogen stages (P<0.05) as well as between anagen and catagen stages (P<0.05), but there was no difference between telogen and catagen stages (P>0.05). HSP90 protein expression was observed in the epidermis, sebaceous gland and hair root sheath in the skin (Fig 11).  In the hair cycle, the highest level of CGI-58 mRNA expression was found during the anagen stage, whereas the lowest expression level was found in the telogen stage. The expression level in the catagen stage was between the anagen and telogen stages. The expression level showed significant difference between anagen and telogen stages (P<0.05) as well as between anagen and catagen stages (P<0.05), but there was no difference between telogen and catagen stages (P>0.05).

Expression of KDF1 in skin during hair cycle
A fluctuation in the relative expression levels of KDF1 mRNA during the hair cycle was shown in Fig 13. In the hair cycle, the highest level of KDF1 mRNA expression was found during the telogen stage, whereas the lowest expression level was found in the anagen stage. The expression level in the catagen stage was between the telogen and anagen stages. The expression level showed significant difference between telogen and anagen stages (P<0.05) as well as between telogen and catagen stages (P<0.05), but there was no difference between catagen and anagen stages (P>0.05).

Modified method
To observe the histological structure of the hair follicle clearly, we used the Sacpic stain method, which was well suited for the visual assessment of follicle activity because it accentuates the inner root sheath. Tissue types were clearly defined. Results: nuclei, dark blue; keratin, yellow; collagen, blue; inner root sheath, bright red; outer root sheath, pale green; smooth muscle, green [42].
The shrinkage effect is the most important problem to prevent when measuring skin thickness. As soon as we harvested the skin, we fixed it to a paperboard by putting pins into the four corners and stored it in 4% paraformaldehyde solution. We believe that this procedure could prevent most of the shrinkage effect in the transverse plane.

Histologic characteristics of skin structure
We confirmed that yak skin is composed of two layers: the epidermis and dermis. The total epidermis of the hairy skin consisted of the stratum corneum and the viable epidermis (stratum basale, stratum spinosum, stratum granulosum). The pelt was composed of compound hair follicles, which produced a primary and some secondary hair follicles. Associated with each primary follicle, there was an arrector pili muscle, a multilobular sebaceous gland, and a coiled tubular sweat gland. The difference between the primary and secondary follicles was that the primary follicles had their own sweat glands but that the secondary follicles did not bear sweat glands [43]. We confirmed that the yak hair follicle group consisted of one primary follicle and several secondary follicles, which was similar to that found in ferrets [19] but different from Iranian sheep breeds [43], Camelus dromedaries [44], Australian cashmere goat [45], llama [17] and sheep [18]. The sweat glands in yaks are not well developed. Sweat secretion does not occur readily, thereby reducing the heat radiation surface. This appears to force the animal to retain heat in the body and helps increase its tolerance to cold [46].

Age-rated thickness change
In this study, we first measured the skin thickness of different ages and in different regions of yak. Skin thickness varied in different regions of the body surface in yak. The thickest-haired skin was present on the cheek, forehead, dorsal neck and ventral neck in the newborn and half-year-old groups. The thinnest part in the newborn and half-year-old groups was the inguinal region. The thickest haired skin was present on the back, followed by the metacarpus and buttocks in the adult group. The thinnest haired skin in the adult group was the inguinal region and the axilla. In a group of newborn and half-year-old yak, the area around the head, cheek and neck were thicker than other parts of the body surface, which was similar to llama [17]. The thickest location was the back in the adults, which corresponds to equine skin [21] and had previously reported in yak [46], perhaps because the back is the part of the body that is most exposed to wind, rain and snow. Yak skin thickness decreased dorsally to ventrally on the trunk. This pattern of skin thickness change was typical of most domestic large animals [47]. In yak, the skin on the lateral surface was thicker than the skin on the medial surface in the limbs.
The total thickness increased with age. Previous researchers noted that sunlight appears to have a considerable effect on the thickness and physical properties of skin [48,49]. Collagen was a major component of skin, and the age-related changes in thickness correlate well with skin collagen content [50]. The significant change in the epidermis in adults was obvious. The corneum layer increased with age, which was similar to the reports of Mugale [51].
Knowledge of skin thickness in yak may be useful in harvesting full-or split-thickness skin grafts to produce leather. Moreover, these results were useful for studying the relation between age-related thickness changes of skin and the living environment.

Expression of HSPs in skin during hair cycle
This study reported for the first time the expression patterns of HSP27, HSP70 and HSP90 in skin during the hair cycle in yak. The HSP27 protein expression in the epidermis suggested that this protein may be useful for keratinocyte cell growth and regeneration, which concurs with previous studies [28][29][30][31][32]. In human epidermal keratinocytes, the expression of HSP27 was closely related to differentiation both in vitro and in situ [52]. HSP27 and p38-MAPK serve essential functions in the maintenance of the epidermal structure, and HSP27 was associated with keratinocyte differentiation [53]. Moreover, the expression of HSP27 in the epidermis showed that HSP27 may be the target for immune response and could protect against pathogens [25,26,[54][55][56].
The highest expression of HSP27 during anagen and its weak expression in catagen and telogen agreed with the results in the mouse model [57]. The expression pattern suggested HSP27 may be involved in the hair follicle cellular cycle [58]. HSP27 expression in the hair cycle could be related to both keratinocyte differentiation and apoptosis in the hair follicle. HSP27 may promote and prolong anagen by protecting hair follicle keratinocytes against apoptosis [59]. Numerous studies have shown that HSP27 inactivates the caspase cascade by binding with caspase-3 and cytochrome C released from mitochondria and that it thus prevents apoptosis [60][61][62]. The weak expression of HSP27 in catagen and telogen may be followed by the process of terminal differentiation and apoptosis of the keratinocyte. HSP27 could mediate this process by inducing some growth factors, such as FGF. In general, the expression pattern suggested that HSP27 expression may correlate with the level of differentiation of the keratinocytes and the level of keratinization of the outer root sheath.
The predominant expression in epidermis of HSP70 and HSP90 protein as well as HSP27 suggested that they may also involve in keratinocyte cell growth and differentiation. HSP70 played an important role in cell apoptosis. In testis, ablation of HSP70 isoform resulted in germ cell apoptosis [63]. HSP90 was known as a molecular chaperone and had other functions. HSP90 can control cell proliferation by stabilizing the client proteins N-RAS and B-RAF [64,65]. The HSP90 protein was weakly expressed in all hair cycle stages compared with HSP27 and HSP70. This result agreed with Wilson's study [66]. Although HSP90 was abundantly expressed in other tissues, it was not largely present in skin [39]. The mRNA expression levels of HSP70 and HSP90 showed no difference among three stages, but both of the two protein expression levels showed significant difference in all three stages, we believed that it related to the process of transcription regulation and would study deeply on this part. Otherwise, it may also be related to the interaction between HSPs proteins.
This study had demonstrated the expression pattern of HSP27, HSP70 and HSP90 in yak skin during hair cycle. All of three HSP proteins were involved in the hair follicle cellular cycle and may related with cell apoptosis. However, the different expression patterns suggested that the function of each HSP protein was various. Our further research at this moment may give a definite mechanism next.

Expression of CGI-58 and KDF1 in skin during hair cycle
We also detected the mRNA expression levels of CGI-58 and KDF1 in skin during the hair cycle in yak. CGI-58 showed the same expression pattern with HSP27 in mRNA level. The expression of CGI-58 mRNA in the anagen stage was the highest, followed by the catagen stage, and the expression in the telogen stage was the lowest. CGI-58 mRNA expression was up-regulated concomitantly with both epidermal stratification and keratinocyte differentiation [67]. The same pattern in skin during the hair cycle in yak suggested both CGI-58 and HSP27 were involved in keratinocyte differentiation in hair follicles. The expression of KDF1 mRNA was contrary to CGI-58.
The highest level was in the telogen stage, followed by the catagen stage, and the expression in the anagen stage was the lowest. KDF1 was expressed in epidermal progenitor cells and the progeny where it curbed proliferation as well as blocked proliferation and promoted differentiation [68]. The cycle-dependent expression of KDF1 suggested it may be relate to the proliferation state of hair follicle keratinocytes.